Author(s)

JiGang Pan, Jie Qian^*, MingLiang Cao

Affiliation(s)

School of Mechanical and Electrical Engineering and Automation, Nanhang Jincheng College, Nanjing 211156, Jiangsu, China. 

Corresponding Author

Jie Qian

ABSTRACT

As the proportion of renewable energy in the power grid continues to increase, accurate day-ahead photovoltaic (PV) power forecasting is critical for ensuring grid stability. This study proposes a systematic framework aimed at addressing four key challenges: a) introducing a two-dimensional bias matrix to quantify seasonal/intraday power fluctuation characteristics; b) establishing a rolling ridge regression model to achieve self-driven forecasting based on historical power data; c) innovatively designing a segmented strategy for sunny/cloudy/rainy weather scenarios to optimise the numerical weather prediction (NWP) fusion process, thereby addressing the variable meteorological impacts in practical applications and significantly improving prediction accuracy under complex meteorological conditions; d) employing weighted interpolation spatial downscaling techniques to refine NWP resolution to the power plant level. Validation results show that downscaling processing improved the Pearson correlation coefficient from 0.64 to 0.76, reduced the Root Mean Square Error (RMSE) from 0.57 kW to 0.46 kW, and decreased the Mean Absolute Error (MAE) from 0.49 kW to 0.34 kW. This integrated solution significantly enhances prediction accuracy, providing robust technical support for grid dispatch in high-penetration renewable energy systems and offering more reliable decision-making basis for smart grid management.

KEYWORDS

Photovoltaic power prediction; Ridge regression; Numerical weather prediction; Spatial downscaling; Scenario segmentation

CITE THIS PAPER

JiGang Pan, Jie Qian, MingLiang Cao. Day-ahead predictions of the power generated by a photovoltaic power plant. World Journal of Engineering Research. 2025, 3(4): 18-27. DOI: https://doi.org/10.61784/wjer3045.

REFERENCES

[1] Sharadga H, Hajimirza S, Balog R S. Time series forecasting of solar power generation for large-scale photovoltaic plants. Renewable Energy, 2020, 150: 797-807.

[2] Luo X, Zhang D, Zhu X. Deep learning-based forecasting of photovoltaic power generation by incorporating domain knowledge. Energy, 2021, 225: 120240.

[3] Wang J, Zhou Y, Li Z. Hour-ahead photovoltaic generation forecasting method based on machine learning and multi objective optimization algorithm. Applied Energy, 2022, 312: 118725.

[4] Radhi S M, Al-Majidi S D, Abbod M F, et al. Machine Learning Approaches for Short-Term Photovoltaic Power Forecasting. Energies, 2024, 17(17): 4301.

[5] Guo X, Mo Y, Yan K. Short-term photovoltaic power forecasting based on historical information and deep learning methods. Sensors, 2022, 22(24): 9630.

[6] Zhang R, Wu Y, Zhang L, et al. A multiscale network with mixed features and extended regional weather forecasts for predicting short-term photovoltaic power. Energy, 2025: 134792.

[7] Seshadri P, TS B P, S S. Time Series-Based Photovoltaic Power Forecasting to Optimize Grid Stability. Electric Power Components and Systems, 2021, 49(16-17): 1379-1388.

[8] Ding S, Li R, Tao Z. A novel adaptive discrete grey model with time-varying parameters for long-term photovoltaic power generation forecasting. Energy Conversion and Management, 2021, 227: 113644.

[9] Ilias L, Sarmas E, Marinakis V, et al. Unsupervised domain adaptation methods for photovoltaic power forecasting. Applied Soft Computing, 2023, 149: 110979.

[10] Zhang X, Yu T, Yang B, et al. A random forest-assisted fast distributed auction-based algorithm for hierarchical coordinated power control in a large-scale PV power plant. IEEE Transactions on Sustainable Energy, 2021, 12(4): 2471-2481.